The problem is that safety means, by definition, ruling out unsafe code at compile time or run time. If you don't prioritise safety early then you run a high risk of discovering when you try to retrofit safety later that you need to rule out a lot of existing code. Even if you haven't promised stability, breaking existing code hurts the ecosystem.

Therefore when setting priorities for language evolution it seems better to identify work that is less likely to result in breaking code, and prioritise safety over that.